Osteoporosis is a degenerative bone disease associated with losses in bone volume and increases in the risk of spontaneous fractures of the spine, hip, and radius. Millions of post-menopausal women are particularly at risk. However, the biomechanical aspects of the disease have received little attention. Thus, it has not been clinically possible to reliably predict which patients are at risk. The current proposal describes the continuation of a three-year grant to study fracture risk prediction in the lumbar spine. In the last funding cycle, we were able to measure vertebral density non-invasively from cadaveric vertebrae and to relate these measures to in-vitro tests of vertebral strength in compression. We showed clinically that predictors based on estimates of the compressive strength of the anterior vertebral centrum provide improved descrimination between fracture and non-fracture groups when compared to available techniques. We are currently completing the analysis of flexion-compression experiments comparing QCT and DPA, and have initiated 3-D finite element modeling studies. The primary objectives of this continuation are to refine fracture risk predictors for the lumbar spine and extend these methodologies to predictions in the thoracic spine. Vertebral density and geometry data obtained from fresh cadaver lumbar spine segments and from QCT measurements will be used in models of lumbar vertebrae. We will consider variations in density, material properties and geometry to examine the interactions between vertebral centrum, cortex, and endplates. The ultimate strength of matched thoracic and lumbar spine segments will be measured using uniaxial compression and combined compression-flexion loading. Quantitative CT measurements will be made for the lumbar and thoracic regions in vitro to determine variations in density along the spine and within the vertebra. The effects of lung tissue on the precision and accuracy of QCT measurements of thoracic vertabrae will also be determined in vitro. Prospective and cross sectional clinical trails will be conducted using the resulting predictors of fracture risk. We plan a five year clinical study a determine the changes in vertebral density associated with clinical interventions in a perimenopausal population. We will measure lumbar vertebral density by QCT annually and apply our fracture risk assessments to controls and women treated with calcium supplement and exercise or estrogen- progesterone. For these subjects we will measure the bone mineral content of the radius and correlate the results with QCT. The eventual goals of these studies are to identify clinical predictors of the osteoporotic state, and to describe the progressive biomechanical changes in the strength of the thoracic and lumbar spine.